The goal of this research is to collect pilot data useful for determining whether the way in which diabetes mellitus (DM) is explained by elderly from different racial/ethnic groups plays a role in their decisions to employ self care. DM is a significant public health problem, especially for minority adults. In collecting this pilot data, we attempt to establish infrastructure for a larger multi-site study of self-care management strategies among racially and ethnically diverse groups of elderly individuals. Specific research questions to be addressed include: Do elderly individuals from differential racial/ethnic groups conceptualize/explain diabetes in the same or different ways? Does race or ethnicity predispose elderly with diabetes to the use of certain self-care management strategies? Among elderly from different racial/ethnic groups, do specific lay explanatory models of diabetes correlate with the use of specific self-care management strategies? For each of four groups, African Americans (Case Western Reserve Univ.), Latinos (University of Toledo), American Indians (Wayne State Univ.), and rural Whites (University of Kentucky), interviews will be conducted with 20 respondents 55 years of age or older identified has having diabetes and 10 individuals 55 years of age or older without diabetes. Study informants will constitute a purposive rather than a statistically representative sample, a design chosen to emphasize the range and variation of responses rather than the representatives of samples. Interviews will include two component parts: (1) A semi-structured interview, including collection of social and demographic data; and (2) self-administration of the multi-factorial Meaning of Illness Questionnaire (MIQ) and two diabetes-specific instruments. The semi-structured interview guide explores symptom experience, causal attributions, and self-care strategies. The MIG has been used in studies of different chronically ill populations, and is here adapted for use with diabetics. Transcriptions of the interviews will be subject to the constant comparative method for data analysis. Factor analysis will be used to specify the factor structure of the MIQ in this population of racially/ethnically diverse elderly adults. A similar strategy will be applied to the diabetes-specific instruments.